Phenotypic Screening – Cardiomyocytes

You are here:

This study shows how we have developed a phenotypic screen using iPSC derived cardiomyocytes for both pharmacological- and toxicological- studies.  

Challenges

There is growing evidence to suggest that cells grown in 3-dimensional (3D) cultures offer a more physiologically relevant model and demonstrate differential pharmacological responses to those grown in 2-dimensions (2D). Refinement of human cardiotoxicity modeling has significant implications for the improvement of compound screens, mechanistic studies, and drug discovery plus development.

Cells grown in vitro are traditionally grown as 2D monolayers, that are easy to set up and maintain. However, these are often monocultures (consisting of only one cell type), lack structural architecture, due to the absence of extracellular matrix, and are flat, with 50% of their surface exposed to tissue culture media rather than being surrounded by other cells. Hence, 2D cell cultures are a poor approximation of tissues in living organisms, particularly elongate striated cells such as muscle cells and more specifically heart muscle cells.

Simple, highly consistent, and easy-to-use 3D models hold great promise to reduce the number of costly drug failures in clinical trials, provide more realistic in vitro models, and generally lead to more useful data and more relevant and successful research. There are a range of techniques used to grow cells in 3D including spheroids and hydrogels. A scaffold is useful when cell-to-cell and cell-to-matrix contacts are important to maintain cell morphology and behavior, for cells that do not in nature form spheroids, where more cells are required for sensitive assays than can be grown in a spheroid, and where diffusion of gases and proteins is desirable.

Electrospun scaffolds provide an easy-to-use, sterile well plate format, compatible with standard laboratory liquid handling techniques and automated image analysis equipment, and require minimal protocol adaption when switching from 2D to 3D. The thickness of the scaffolds is also important to facilitate microscopic imaging in situ, while providing the benefits of 3D cell culture to observe 3D cell morphology and behavior. By providing an aligned scaffold with highly consistent architecture, this leads to reproducible results in cell based assays well-to-well and batch-to-batch, minimizing experimental variability.

Figure 1. Microscopy images of electropsun fibre as a random fibre layer (left) and aligned fibres (right). Aligned fibres allow cells with elongated shape and polarity to adopt a more physiological geometry.

Inducible Pluripotent Stem Cells (iPSC’s) are available from several commercial suppliers derived from different human patient populations. These iPSC’s are driven down different differentiation paths in vitro by the providers to obtain differentiated cells for use in screening. These can differ in quality and quantity. 

In this study, we have evaluated three different sources of human iPSC derived cardiomyocytes: two commercially available (Cellular Dynamics International “iCell”, and Axiogenesis “Cor4U”) and one non-commercial source. Cells were grown in either a classic 2D culture or on 3D aligned electrospun nanofibre plates and treated with a range of pharmacological agents representing different targets or modes of action. 

The FLIPR Tetra system was used to measure the potential of several pharmacological agents to modulate calcium mobilisation in 2D and 3D environments. Specifically, we analysed beat count & frequency, amplitude of contractions and peak rise and decay to quantify changes in the phenotypic profile.

What We Did...

FLIPR Tetra Screening

Cardiomyocytes were selected from current manufacturers of iPSC-derived cardiomyocytes. All cells were prepared and plated as per the manufacturer’s recommended protocols on either standard tissue culture (2D) or 3D, 384 well aligned electrospun nanofiber plates (Nanofiber Solutions). All cells were monitored for synchronous beating prior to treatment. 

To measure changes in the phenotypic profile of the cardiomyocytes, we utilized the FLIPR Tetra with the EarlyTox cardiotoxicity kit (Molecular Devices). This assay is based on a calcium sensing dye that monitors calcium flux (mobilization) during cardiomyocyte beating. All work on the FLIPR was performed in a 384 well 2D or 3D aligned fiber plate for screening purposes.

Compound

Structure

Target and Mode of Action

Phenotypic Effect

Isoproterenol

ß-1 and ß-2 adrenoceptor antagonist

Increase beat rate (tachycardia)

Propanolol

Cardiomyoctes_Phenotypic screening_Propanolol

ß adrenoceptor antagonist

Decrease beat rate (bradycardia)

Sotalol

Cardiomyoctes_Phenotypic screening_Sotalol

ß adrenergic receptor antagonist

Decrease beat rate (bradycardia) Inhibit the efflux of potassium ions that delays electrical signaling decreasing beat rate

Astemizole

Cardiomyoctes_Phenotypic screening_Astemizole

Histamine H2 antagonist/hERG channel blocker

QTc interval prolongation

2-APB

Cardiomyoctes_Phenotypic screening_2-APB

IP3 receptor antagonist

Concentration dependent effect: Stimulates store operated calcium release at low concentrations and inhibits at high concentrations

Table 1. Pharmacological compounds used to influence cardiomyocyte phenotypic behavior.

After an initial baseline reading, all cardiomyocytes were treated in a dose-dependent manner with a range of pharmacological compounds known to affect the cellular phenotype. Compounds were prepared and dispensed simultaneously to all wells in the FLIPR Tetra instrument. 

Plates were then monitored (for 60 seconds) every 15 minutes for up to 2 hours post-compound addition. All data was acquired and analysed using FLIPR Screenworks Peak Pro software.

Comparing Baseline Parameters of 2D and 3D Cardiomyoctes Models

Cardiomyocytes from different manufacturers showed significant differences in baseline beats per minute (BPM) and amplitude in 2D and 3D culture (Figure 2). Human cardiomyocytes were sourced from Cellular Dynamics International (CDi), Axiogenesis and a non- commercial partner. All cells were grown in classic 2D tissue culture or on electrospun nanofiber plates (3D) prior to measurements on the FLIPR Tetra. We observed that for all cells, growth in 3D resulted in increased baseline BPM and a reduced beat wave amplitude (Figure 2). Analysis of FLIPR wave profiles (Figure 3) indicates that the beat pattern of 2D and 3D are very different and 3D is more cyclical in nature. This may have important implications for compound responses, mechanistic studies and assay screening when selecting a source of cardiomyocytes for cell based assays.

Cardiomyoctes_Phenotypic screening_beat rate

Figure 2. Beat rate (left) and amplitude (right) for cardiomyocytes from different sources grown in 2D and 3D (electrospun nanofiber plates). Although beats per minute was consistent between 2D and 3D, the amplitude of beats was different in 2D versus 3D (BPM +/- SD n = 3).

Cardiomyoctes_Phenotypic screening_beat frequency

Figure 3. Beat frequency and amplitude patterns of cardiomyocytes from different sources in 2D and 3D. Cells in 3-D have a more cyclical beat pattern (representative well data shown).

Increased Sensitivity of Phenotypic Assays Using 3D Culture of Human IPSC Derived Cardiomyocytes

CDi iCell cardiomyocytes show increased sensitivity and observable phenotypic difference in response to compounds on 3D nanofiber plates. Cardiomyocytes were treated with sotalol or propranolol and measured every 15 minutes for 45 minutes. 

Sotalol treated cells in 3D showed a decrease in beats per minute in a concentration dependent manner, compared to 2D (Figure 4). The effect of sotalol was transient and the cells recovered to baseline levels after 45 minutes (Figure 4). 

In contrast, propranolol treated cells did not recover after treatment (Figure 5). However, we did observe a significant increase in assay sensitivity and range in 3-D nanofiber grown cells (Figure 5).

Cardiomyoctes_Phenotypic screening_sotalol treated cells

Figure 4. Sotalol treated CDi cells demonstrated a difference in the amplitude and shape of the calcium response curve with a more protracted recovery period following sotalol treatment, combined with a transient reduction in beats per minute (BPM) at higher concentrations at 15 minutes post-treatment (BPM +/- SD n = 3).

Figure 5. Propranolol treated CDi cells demonstrated a similar shape of the calcium response curve, but different beat frequency, coupled with significant dose- and time- related decreases in 3D compared with 2D (BPM +/- SD n = 3).

Dose Response to Pharmacological Agents - Astemizole

Cardiomyocyte response to astemizole varied depending on the source of cells and the growth conditions. Human iPSC-derived cardiomyocytes were either grown in a standard tissue culture plate (2D; top panel) or electrospun nanofiber plate (3D; bottom panel) and subsequently treated with various concentrations of astemizole. Data expressed as beats per minute (BPM +/- SD n = 3).

Cardiomyoctes_Phenotypic screening_changes in beat rate

Figure 6. Changes in beat rate of cardiomyocytes from different commercial suppliers cultured in 2D and 3D in the absence (blue) or presence (red) of different concentrations of astemizole 15 minutes post-treatment (BPM +/- SD n = 3). Note high concentrations ceased the beating of the cardiomyocytes.

Overall, cardiomyocytes responded in a dose-dependent manner with increasing concentrations of astemizole having a negative effect on the BPM of all cardiomyocytes tested, compared to untreated controls. The non-commercial cells proved to be more sensitive to astemizole treatment and toxic at the higher concentrations, completely inhibiting the beating process. For both CDi and Axiogenesis derived cells, growth in 3D improved the sensitivity and dynamic range of the assay compared to cardiomyocytes in 2D conditions. 

Summary

In conclusion, there was significant variation between the resting BPM and beat amplitude in 2D iPSC-derived cardiomyocytes compared to 3D aligned fiber cells. 3D cardiomyocytes were more effective for observing the potentially subtle changes in beat peak profile caused by pharmacological active compounds.

These observations may be due to the unique ability to culture the cells on aligned fibers, resulting in the cells forming a polarity along the length of the fiber, then connecting laterally between fibers. Observations suggest that all cells in each well beat to the same frequency, unlike those seen in spheroids in which two beat frequency sources can often be observed, making it difficult to record beat frequency and impossible to measure amplitude.

Within the plate, although cells in each well beat to the same frequency, the cells in adjacent wells beat at other frequencies dictated by the cell population within individual wells. This assay was also amenable to 384 well format, making it ideal for screening purposes.